An Experiment on Automated Requirements Mapping Using Deep Learning Methods
| Autor: | Felix Petcusin, Liana Stanescu, Costin Badica |
|---|---|
| Rok vydání: | 2019 |
| Předmět: |
Requirements engineering
business.industry Integration testing Computer science Deep learning Time to market Automotive industry 020207 software engineering 02 engineering and technology Software 0202 electrical engineering electronic engineering information engineering 020201 artificial intelligence & image processing Artificial intelligence business Software engineering Quality assurance Natural language |
| Zdroj: | Intelligent Distributed Computing XIII ISBN: 9783030322571 IDC |
| DOI: | 10.1007/978-3-030-32258-8_10 |
| Popis: | Requirements engineering is one of critical activities in systems development for Automotive Industry. Its outcome is most often represented by a set of documents capturing requirements specifications in natural language. For quality assurance and maturity support of the final products, the requirements must be verified and validated at different testing levels. To achieve this, the requirements are manually labelled to indicate the corresponding testing level. The number of requirements can vary from few hundreds in smaller projects to several thousands in larger projects. Their manual labeling is time consuming and error-prone, thus sometimes incurring an unacceptable high cost. In this paper we report our initial results on the automated classification of requirements in two classes: “Integration Test” and “Software Test” using Machine Learning approaches. Our solution could help the requirements engineers by speeding up the classification of requirements and thus reducing the time to market of final products. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|